커밋 그래프

2282 커밋

작성자 SHA1 메시지 날짜
Jacob Lambert 2bd527c676 SWDEV-387063 - Use clang default for C++ version
Instead of enforcing c++14 here, we can instead use the current
clang default

Change-Id: Ib0a178a53c1377f2910edf6fab82b2bac6567ac7


[ROCm/clr commit: 33e48b9629]
2025-02-03 11:07:52 -05:00
Anusha GodavarthySurya 837f7ca08c SWDEV-489084 - Avoid creating internal stream when graph has single branch
Change-Id: I9371d44481257069bb51c0217a57f97d803589c4


[ROCm/clr commit: b385992f94]
2025-01-31 00:16:57 -05:00
Saleel Kudchadker 21ae9ef25e SWDEV-508225 - Improve fat binary handling
Change-Id: I78a9951f2f4c4c743c1205b1e40aac215054e27d


[ROCm/clr commit: 08af3eb484]
2025-01-28 14:38:21 -05:00
Marko Arandjelovic 8b956db8db SWDEV-489619 - Added checks for memcpy capture path
Change-Id: I0e156099282f0b6393bcbcee2e9b96c31034a851


[ROCm/clr commit: 269ec54252]
2025-01-27 03:51:34 -05:00
Ioannis Assiouras 8169855390 SWDEV-510319 - Fixed random segfaults in graph tests
This change fixes random segfaults in graph tests that
are seen after the change make internal callbacks non-blocking.
The callback thread that decreases the GraphExec ref count
may now run after the runtime shutdown. This can cause a segfault
because the hip::device that is accessed in GraphExec destructor
is already destroyed during runtime shutdown. This patch ensures
that the hip::device object  stays alive until after the
callback thread completes.

Change-Id: I75a6ac01f27a0b2250bbd10ed389ebfb322927af


[ROCm/clr commit: 21c223f8df]
2025-01-25 09:54:15 -05:00
Sourabh Betigeri e3c4a81b69 SWDEV-502219 - Adds validity checks for negative parameters passed
Change-Id: Ib8a531533306a27143d74b81c074de81051eb896


[ROCm/clr commit: c460b0541b]
2025-01-24 16:32:29 -05:00
zichguan-amd 4e84f2182e SWDEV-509518 - Allow LLVM_ROOT and Clang_ROOT to be used with find_program
Fixes #123. find_program doesn't follow CMP0074 and thus ignores LLVM_ROOT and Clang_ROOT. This change adds LLVM_ROOT and Clang_ROOT to the search path of find_program for llvm-mc and clang in hiprtc to mimics previous add_package behaviour.
Caveat: cmake-specific variables like CMAKE_PREFIX_PATH will take precedence over paths specified with HINTS for find_program, there's no way to change the ordering unless we skip cmake-specific variables all together using NO_CMAKE_PATH and NO_CMAKE_ENVIRONMENT_PATH.

Change-Id: I1fedb60cda09744416e19b3c6e3e0c5c9045f8e7


[ROCm/clr commit: 272ef9a7bf]
2025-01-23 11:50:36 -05:00
Jaydeep Patel f674ba58f0 SWDEV-508982 - [6.4 Preview] - Handle hipMemPoolCreate, hipMemPoolDestory & hipDeviceSetMemPool during stream capture.
Change-Id: Ia195442041803896df814798c3d2053c0ba7770c


[ROCm/clr commit: 57df1b348f]
2025-01-22 05:28:47 -05:00
Jaydeep Patel 1d7b7cde76 SWDEV-457316 - Use phy memory obj stored in user data instead of querying from memObjs.
Change-Id: Id837eb00195d88b50904441f01cf8153fa752ecd


[ROCm/clr commit: b4df9fb6ec]
2025-01-21 22:05:14 -05:00
Jatin Chaudhary 0795f00a14 SWDEV-341217 - Initial work to use SPIRV in HIP
Change-Id: If5c09b5e86b498e7ac5eb05adf28cb7a1fac8101


[ROCm/clr commit: 6a5d19059d]
2025-01-20 03:54:23 -05:00
Vladana Stojiljkovic 4b786b5207 SWDEV-498061 - Add capture support for hipModuleLaunchCooperativeKernel
Change-Id: I5ed188e046c680c2785b3952391f59ed1d0c21b8


[ROCm/clr commit: 30cb2d0e67]
2025-01-16 10:54:30 -05:00
Marko Arandjelovic cf8eeabfe2 SWDEV-489619 - Fix memcpy tests with capture stream enabled
- Added missing validation as graph node should not be created
 if parameters are invalid
 - Fix conversion of input params to graphNode params

Change-Id: I37ab04942b5fb2eb07386850cb7dbbf26f9ca967


[ROCm/clr commit: db8527f655]
2025-01-16 10:31:04 -05:00
Marko Arandjelovic 8647bb483b SWDEV-504084 - Make hipModuleGetFunction use the device the module is loaded on
If a module is loaded on one device, hipModuleGetFunction and other similar APIs should be able to run successfully from another device.

Change-Id: I96084cbd6c6dcf2a81019779a6ab1842ef2f35d1


[ROCm/clr commit: c46f843b99]
2025-01-16 10:16:42 -05:00
Ioannis Assiouras 67c93c3bad SWDEV-505503 - Use internal device synchronize function in __hipUnregisterFatBinary
This is to avoid calling the HIP_INIT macro during the shutdown process.

Change-Id: I2e65f6e10491918a17445ee1e8ddd08286070358


[ROCm/clr commit: 5e3a29078d]
2025-01-15 18:57:34 -05:00
Sourabh Betigeri c95bb38b2b SWDEV-507960 - Return with error code if stream of type hipStreamLegacy is being attempted to destroy
Change-Id: Iee7ada6a5a905b44360a7e4049fc8b1a45c80db0


[ROCm/clr commit: 9d8d35ae40]
2025-01-13 18:17:37 -05:00
Ioannis Assiouras 161d590405 SWDEV-503760 - Only consider allocations that are less than X% larger in a mempool request
Change-Id: I94acbca606fd4c575e2e1a9e34959ce650571867


[ROCm/clr commit: 44b6b6813d]
2025-01-13 16:57:26 -05:00
Rahul Manocha 0a9315f5e0 SWDEV-497288 - Add validation checks to hipGraphExecNodeSetParams
Change-Id: If8cde47bc8e62414333768e01064298d8a3d80ee


[ROCm/clr commit: 2b32d9aada]
2025-01-13 16:49:45 -05:00
Anusha GodavarthySurya 08c92f4793 SWDEV-480209 - Make internal callbacks non-blocking
Change-Id: Ic918d08f341abfd9a7c167d09f9c723cdc43157f


[ROCm/clr commit: 683a942364]
2025-01-10 02:16:11 -05:00
Saleel Kudchadker 6a416008aa SWDEV-508004 - Improve hipEventRecord
- Resolve stream once for event record. We should avoid calling
  getStream again in addMarker

Change-Id: I78448c4f151ae10a5c8e8c248b2f4078b84191cb


[ROCm/clr commit: a22c45d635]
2025-01-09 16:47:46 -05:00
German Andryeyev 5a76761960 SWDEV-507019 - Change the function lock to the module lock
Multiple functions can be located in the same module.

Change-Id: Ia4ca3db64fe5b0822584059d3770c91103665c63


[ROCm/clr commit: 45a12208b6]
2025-01-09 12:35:29 -05:00
Rahul Manocha b1ef5972d6 SWDEV-504215 - fix rocalution perf drop by disabling cpu wait
Change-Id: I878f3420073b05cc6241f524ac428e47c0ce823d


[ROCm/clr commit: 05baf9ff22]
2025-01-07 17:02:24 -05:00
Marko Arandjelovic 44ff3ba1cc SWDEV-506234 - Refactor validation in hip_memory
Change-Id: I9d69695e4b6668e6de00f1f6b060862872358340


[ROCm/clr commit: 7e152bb0f3]
2024-12-31 00:35:25 +02:00
Jaydeep Patel f20f399915 SWDEV-505276 - Parent graph of orig node and original graph of cloned node should be same.
Change-Id: I6ebc21cc42e41ad5d952a69fb3b3cb095f32cffb


[ROCm/clr commit: dff8197b1d]
2024-12-24 04:32:14 -05:00
taosang2 da613fbbeb SWDEV-479958 - Support different address mode
Support different address modes in X, Y, Z directions

Change-Id: If1db5a8af33c92dd14b48968c3e8eceb97daea6c


[ROCm/clr commit: d82d6a78cf]
2024-12-23 16:39:54 -05:00
Ioannis Assiouras b4019892c9 SWDEV-505504 - Disable vectorization in GetHipDispatchTable
Change-Id: Id33144623555a5d25e029ca644f6274610dcd0ad


[ROCm/clr commit: 158b6a29e0]
2024-12-20 17:47:07 -05:00
Jaydeep Patel 12ed697705 SWDEV-505205 - Fix hipStreamLegacy segfault with hipStreamWaitEvent.
Change-Id: I17fdaf7ac323507f99a7c071066944296537489c


[ROCm/clr commit: a05a02e527]
2024-12-20 04:18:21 -05:00
German Andryeyev 55d4a75016 SWDEV-504658 - Reduce the lock scope for kernel look-up
The vector with all kernels is preallocated on the executable init.
Thus, reduce the scope  of global lock to the binary creation only.

Change-Id: I73035013a6562175069137e895bba815f466ee35


[ROCm/clr commit: 0640d36019]
2024-12-18 17:04:51 -05:00
Sourabh Betigeri 02c203bf9c SWDEV-505277 - Adds hipStreamBatchMemOp in the enum of hcc_map
Change-Id: I6e58dfbe4ba13db8717edc36020fefabc9ddbe23


[ROCm/clr commit: cd9db5a2fa]
2024-12-18 05:38:58 -05:00
German Andryeyev d5b3b0830a SWDEV-504650 - Switch to shared_mutex for events
Use shared mutex for events validation

Change-Id: Iff291c758d9edd65717c506150f3b9d39e5306ba


[ROCm/clr commit: e3efce20be]
2024-12-17 11:04:58 -05:00
Ioannis Assiouras 2c8805e536 SWDEV-483134 - Remove hipExtHostAlloc API
Change-Id: I60777ef5c56b60dd8100d0d794ca10fb3b96a555


[ROCm/clr commit: e8b2fdab96]
2024-12-16 17:13:49 -05:00
Anusha GodavarthySurya 4288640f69 SWDEV-469422 - Derive GraphExec from Graph and ChildGraphNode from GraphExec
Change-Id: I54d67a1665355579bc249d8ff4f9806e9ee14588


[ROCm/clr commit: 13e2e797c0]
2024-12-16 00:43:57 -05:00
Sourabh Betigeri 750313dfed SWDEV-421020 - Adds hipGraphAddBatchMemOp, SetGetParams and execSetParams APIs
Change-Id: Ieccecfe6173cc68fd3c01f86c99f7cc09fe194a3


[ROCm/clr commit: f1c05e9026]
2024-12-13 06:23:39 +00:00
Sourabh Betigeri 7261404002 SWDEV-440866 - [hip-roclr] Adds support to batch memory operations APIs
Change-Id: I5ac63a6626af8c2b4ac382c52dfe1aaf0b3716b8


[ROCm/clr commit: 03dbcd8ca7]
2024-12-12 19:29:24 -05:00
Tao Sang fb76b9620c SWDEV-496667 - Support gfx9-4-generic target
Support gfx9-4-generic target to cover mi3XX.
Support features sramecc and xnack in generic target.
Improve some code formats.
Add more log on compiler.

Change-Id: I6b3c6af55c60cffd43ce6f17b75998f751b75713


[ROCm/clr commit: 3ad8f1b811]
2024-12-12 14:43:39 -05:00
Saleel Kudchadker 52a3751389 SWDEV-301667 - Improve logging for hip_memory
Change-Id: Id624b2c91e6b701bc0ee561a0c193f2c66654890


[ROCm/clr commit: 537f2fffc9]
2024-12-12 14:42:00 -05:00
Anusha GodavarthySurya acfcd1098f SWDEV-469422 - Refacor childgraph node
Remove static functions in graph

Change-Id: I4df94915f81f250acaea60398aea32ef0ed658e2


[ROCm/clr commit: 28cbf2bc4f]
2024-12-12 12:38:24 -05:00
Jaydeep Patel 704be8ab01 SWDEV-496544 - Sync with stream if it is different than srcMemory stream.
Change-Id: I9c0f94a9531555278a51202ec7203961e1344c2e


[ROCm/clr commit: c8afd7109d]
2024-12-12 06:17:39 -05:00
Branislav Brzak 92b1136755 SWDEV-490860 - Do signal_is_required detection post graph schedule
Change-Id: Iaf1067a811aeac3d16c08de954036e219b545e07


[ROCm/clr commit: 89dfdc4dbe]
2024-12-09 03:57:44 -05:00
Jaydeep Patel a7406bbd5d SWDEV-502532 - Exit graph launch in case of empty graph.
Change-Id: Ifb6ab14ca6810cbc1c9e38c59d1d9e7d367358d9


[ROCm/clr commit: 0d4823ff88]
2024-12-07 12:27:53 -05:00
Ioannis Assiouras dc3ca8aab7 SWDEV-497759 - Fix memObj offset computation for hipHostRegister on Windows mgpu
On Windows, hipHostRegister may add a single object in the MemObjMap
that maps to memory that is allocated on different devices.
This change ensures that the offset that is returned from
getMemoryObject() is computed relative to the memory that is allocated
on the current device.

Change-Id: I5fd3af200bf6f4926fdeaea12dcb9d0154d3a843


[ROCm/clr commit: e80442fdbf]
2024-12-05 16:18:10 -05:00
Marko Arandjelovic 12dc02b4f8 SWDEV-495609 - Change include path for rocclr/utils
- Header files inside rocclr/utils when included from hipamd or opencl should be included as #include "rocclr/utils/xxx.h" instead of "utils/xxx.h"

Change-Id: Ic0760c33b9d091f5620dec67e5482c9698d22093


[ROCm/clr commit: 78f62d3230]
2024-12-05 11:44:20 -05:00
Branislav Brzak cf25aaddcf SWDEV-490860 - Include signal_is_required in dot file dumps
Change-Id: Iec4b433b11fbecb71a4ce68beb7d6f681d25b8e6


[ROCm/clr commit: f2dba978f5]
2024-12-05 04:51:59 -05:00
Shane Xiao dabe311bd8 SWDEV-492049 - Remove the handle of Phy Mem from Memobj
The hipGraph will use VMM by default when allocating memory.
However, the handle of Phy mem has been added to Memobj by default.
Since the Memobj will track the whole address range from handle to
handle + size, this needs the system to reserve the whole address
range. If the system range have not reserved by the system, then it
will have the potential issue that clr finds the Memobj incorrectly.

This patch removes the handle from the Memobj to fix this potential
issue.

Change-Id: I2da38e6b2d11d0d48e1afe66c46899500c290624


[ROCm/clr commit: 231b2410a0]
2024-12-04 19:39:52 -05:00
German Andryeyev 6933aa7c29 SWDEV-501403 - Switch to std::shared_mutex for streamSetLock
Shared mutex allows to have access to the list of streams  from
multiple threads at the same time.

Change-Id: Ibee64b846cde03321d5b17dbee2829c0bab7e7d6


[ROCm/clr commit: efd3ea4b30]
2024-12-04 12:06:51 -05:00
Anusha GodavarthySurya b383c1e443 SWDEV-469422 - Always schedule graph nodes
Change-Id: Icc636527fa19e7bf3eb111bc4b1bb9a5f9acff73


[ROCm/clr commit: b89977d518]
2024-12-03 23:44:23 -05:00
Saleel Kudchadker 957b460014 SWDEV-494149 - Improve hipGet/Set Device
Change-Id: If8975687a3ba9caadafc48a0066f19a4ebaab9e2


[ROCm/clr commit: 6611cc015d]
2024-12-03 13:36:38 -05:00
Sourabh Betigeri 1712acdd2e Revert "SWDEV-440866 - [hip-roclr] Adds support to batch memory operations APIs"
This reverts commit ab0ff9163d.

Reason for revert: hipInfo fails on windows. Updating llvm amd-mainline-closed

Change-Id: I57e1fa1945188b0bc0a799c4f3d540f2b7713003


[ROCm/clr commit: 2ca644cf22]
2024-12-02 16:46:12 -05:00
Marko Arandjelovic 36ba236426 SWDEV-499794 - Update AQL packet after updating GraphNode
Change-Id: I332d70bdf42a276894a548a02d636e370c2ca08c


[ROCm/clr commit: 08aee16573]
2024-12-02 12:29:35 -05:00
Sourabh Betigeri ab0ff9163d SWDEV-440866 - [hip-roclr] Adds support to batch memory operations APIs
Change-Id: I449ffca44bbb04d13348d112e896d603c70fd485


[ROCm/clr commit: bd5d8e9baf]
2024-11-30 17:54:32 -05:00
Anusha GodavarthySurya a3c4a5a19c SWDEV-469422 - Cleanup graph code remove parallellists and nodewaitlists
Change-Id: I00c7b2894333bd13d47b913d3fcdd6e1ffcb741f


[ROCm/clr commit: c47f9dda58]
2024-11-30 04:40:51 -05:00